Audiologic Management of Individuals Receiving Cochleotoxic Drug Therapy

Introduction

Any drug with the potential to cause toxic reactions to structures of the inner ear, including the cochlea, vestibule, semicircular canals, and otoliths, is considered ototoxic (Govaerts et al., 1990; Miller, 1985). Drug-induced damage affecting the auditory and vestibular systems can be called, respectively, cochleotoxicity and vestibulotoxicity. Although ototoxicity can result from occupational and/or environmental exposure to ototoxins, the majority of cases result from drug therapy (Brummett, 1980). The benefits of ototoxic drugs must be weighed against their potential for permanent damage to the inner ear. In designing treatment protocols, the health care team should consider the roles of hearing and balance in maintaining quality of life following therapy.

Permanent hearing loss or balance disorders caused by ototoxic drugs can have serious vocational, educational, and social consequences. These effects may be minimized, or even prevented, if the ototoxic process is detected early during treatment. An effective monitoring program should detect ototoxic damage before the patient becomes aware of ototoxic symptoms. Such early detection allows health care providers to consider treatment alternatives such as modifying drug dosage or changing to a less toxic drug to slow or halt the progression of inner ear damage. Unfortunately, monitoring for ototoxicity is not a common practice and measurement procedures tend to be inconsistent, largely because guidelines specifying patient selection, monitoring methods, and criteria for interpreting results do not exist (Esterhal, Bednar, & Kimmelman, 1986; Meyerhoff, Malle, Yellin, & Roland, 1989).

The ASHA Scope of Practice for audiologists clearly includes the assessment of both auditory and vestibular function (ASHA, 1990). The fact that some medications may selectively affect the vestibular end organ also identifies the area of vestibulotoxicity monitoring as a direct concern of audiologists. For the sake of completeness, the relative potential for cochleo-versus vestibulotoxic effects of known or suspected ototoxic agents is described herein. However, at this time there are no generally accepted vestibulotoxicity monitoring protocols that are efficient, reliable, and completely suitable for application with ill patients. Further research in this area is needed before specific vestibulotoxicity monitoring recommendations can be made. This document, therefore, will concentrate on providing guidelines for the audiologist with respect to serial assessment of cochleotoxicity in patients receiving treatment with potentially ototoxic drugs.

Return to Top

Role/Responsibilities of the Audiologist

Although the responsibility for designing and implementing an auditory monitoring program for cochleotoxicity rests with the audiologist, the implementation and continuation of such a program requires a collaborative effort between audiologists and medical center personnel. The process of identifying patients who are potentially at risk for incurring cochleotoxicity should be initiated by the audiologist in concert with appropriate medical center personnel. The interaction between physicians and audiologists is valuable with respect to decision processes that weigh the need for ototoxic medications against the risk of acquired auditory and vestibular pathology. The audiologist is also responsible for implementing appropriate rehabilitation measures as necessary.

It may be necessary for audiologists to coordinate in-service programs with appropriate medical center personnel. These programs can be helpful to the success of monitoring programs by presenting basic information on the effects that patients may experience from ototoxic drugs and the potentially interactive effects of age, noise, physical condition, and prior drug exposure. A concerted effort by professional organizations in advocating monitoring programs will also improve the acceptance and enhance the success of such programs.

Return to Top

Ototoxic Drugs

Approximately 200 drugs have been labeled ototoxic (see list in Govaerts et al., 1990; Lien, Lipsett, & Lien, 1983; Rebuke, 1986). Different ototoxic drugs can cause either permanent or temporary structural damage of varying degree and reversibility (Bendush, 1982; Brummett, 1980). Those of greatest concern for permanent effects are the aminoglycoside antibiotics and the cancer chemotherapeutic agent cisplatin (and possibly its analogue, carboplatin). The ototoxic effects of salicylate analgesics, quinine, and loop diuretics are usually temporary (although loop diuretics can potentiate the ototoxicity of other drugs, as discussed below; Brummett, 1980). Cellular mechanisms of ototoxicity, well-described elsewhere (De Groot, Huizing, & Veldman, 1991; Govaerts et al., 1990; Laurell & Bagger-Sjoback, 1991; Lerner, Matz, & Hawkins, 1981; McAlpine & Johnstone, 1990; Miller, 1985; Rybak, 1986), will not be discussed in this document.

Aminoglycosides. Since the introduction of streptomycin in the 1940s (Schatz, Bugie, & Waksman, 1944), aminoglycosides have attained widespread use. It has been estimated that about four million patients annually are potentially at risk for hearing loss associated with aminoglycosides in the United States each year (Kumin, 1980). Of all ototoxic drugs, the aminoglycosides are the most vestibulotoxic (Cass, 1991), although they vary greatly in their differential effects on vestibular and cochlear structures (Dulon, Aran, Zajic, & Schacht, 1986). Kanamycin, amikacin, neomycin, and dihydrostreptomycin are preferentially cochleotoxic. Gentamicin affects both cochlear and vestibular structures. Many cases of gentamicin-induced vestibular toxicity have been reported, however, with no apparent concomitant hearing loss (although asymptomatic or undetected high-frequency loss may have occurred). Vestibulotoxic aminoglycosides, in order of decreasing severity, are thought to be streptomycin, gentamicin, tobramycin, and netilmicin (Govaerts et al., 1990). Kanamycin and gentamicin appear to be approximately equal in cochleotoxicity, followed by tobramycin, then netilmicin.

Because aminoglycosides are cleared more slowly from inner ear fluids than from serum (Federspil, 1981), there may be a significant latency to the ototoxic effect of aminoglycosides. This latency can result in initial hearing loss, or progression of loss, following cessation of treatment (Gatell et al., 1984; Lerner & Matz, 1979; Meyerhoff et al., 1989).

Cancer Chemotherapeutics. Although very effective in treating a variety of cancers in adults and children, the chemotherapeutic agent cisplatin is highly ototoxic (Aguilar-Markulis, Beckley, Priore, & Mettlin, 1981; Boheim & Bichler, 1985; Laurell & Borg, 1988; Laurell & Jungnelius, 1990; Pasic & Dobie, 1991; Skinner, Pearson, Amineddine, Mathias, & Craft, 1990; Waters, Ahmad, Katsarkas, Stanimir, & McKay, 1991) and may be the most potent of known ototoxic drugs (Barr-Hamilton, Matheson, & Keay, 1991). It produces irreversible effects, seen initially in the high-frequency region of the inner ear (Fausti et al., 1993; Fausti, Schechter, Rappaport, Frey, & Mass, 1984b; Pollera et al., 1988). Although chemically and functionally quite different, cisplatin and the aminoglycosides cause similar cochlear pathology and symptomatic hearing loss (Brummett, 1981). Vestibulotoxicity from cisplatin is uncommon (Schweitzer, Rarey, Dolan, Abrams, & Sheridan, 1986), but is occasionally seen with very high doses (Cass, 1991).

Ototoxicity of cisplatin as a single therapeutic agent is well documented. Less is known, however, about its interaction with other drugs and/or cranial irradiation. For example, ifosfamide, given as a single agent, is not known to be ototoxic but appears to potentiate the ototoxicity of cisplatin (McHaney et al., 1992). Cranial irradiation alone appears to have little effect on hearing, but prior cranial irradiation appears to exacerbate cisplatin's ototoxicity (Grunberg, Bertram, McDermed, & Apuzzo, 1987; Khan et al., 1982; McHaney, Thibadoux, Hayes, and Green, 1983; McHaney et al., 1992; Schell et al., 1989).

Carboplatin, an analogue of cisplatin, was developed as a less toxic alternative to its parent drug. It has been used for treating ovarian cancer (Adams et al, 1989), a variety of rare or treatment-resistant pediatric tumors (Wake, Takeno, Ibrahim, Harrison, & Mount, 1993) and brain tumors in adults and children (Brummett, Guitjens, Vestergaard, & Johnson, 1993). Carboplatin was initially reported to be non-ototoxic (Calvert et al., 1982), but adverse effects on hearing have since been documented (Kennedy, Fitzharris, Colls, & Atkinson, 1990; Wake et al., 1993). It is, however, considerably less ototoxic than cisplatin (Brummett et al., 1993).

Loop-Inhibiting Diuretics. All diuretics that inhibit the ascending Henle loop in the kidney (loop diuretics) have been found to alter cochlear function (Brummett, 1980; Elidan, Lin, & Honrubia, 1986). The effects of loop diuretics on the vestibular system are controversial (Elidan et al., 1986). Furosemide, bumetamide, and ethacrynic acid produce primarily temporary hearing losses (Matz, 1976; Pillay, Schwartz, Aimi, & Kark, 1969; Schneider & Becker, 1966; Schwartz, David, Riggio, Stenzel, & Rubin, 1970; Silverstein & Begin, 1974). The concurrent administration of loop diuretics and aminoglycosides produces synergistic cochleotoxicity (Brummett, 1981; Brummett, Traynor, & Brown, 1975; Hoffman, Whitworth, Jones, & Rybak, 1987; West, Brummett, & Himes, 1973). Synergism also occurs when loop diuretics are administered with the nonaminoglycoside antibiotics viomycin, capreomicin, fortimicin, and polymyxin B (Davis & Brummett, 1979). The synergistic interaction, however, does not seem to occur with non-loop-inhibiting diuretics such as mannitol, mercuhydrin, or hydrochlorthiazide (Brummett, 1981). There is evidence that the loop diuretics ethacrynic acid, furosemide, and bumetanide also augment cisplatin cochleotoxicity (Brummett, Fox, Russel, & Davis, 1981). In animal studies, the cochleotoxicity of carboplatin was greatly increased by coadministration with furosemide (Brummett et al., 1993). It seems clear that the administration of loop diuretics in conjunction with almost any other drug should be carefully observed for ototoxic potentiation (Brummett et al., 1981).

Salicylate Analgesics. Salicylates are ototoxic and, because they are available over the counter, a large population risk ototoxicity from their use (Miller, 1985). Commonly used salicylates are acetylsalicylic acid (aspirin) and sodium salicylate. Salicylate ototoxicity is most often manifested as high-pitched or hissing tinnitus and hearing loss, with occasional vestibular dysfunction (Cass, 1991). The great majority of reported cases of salicylate-induced hearing loss have been bilateral, symmetrical across frequencies, and reversible (Miller, 1985), although permanent hearing loss has occurred (Jarvis, 1966; Kapur, 1965). Partial recovery of hearing function is usually noticeable within 24–48 hours after salicylate ingestion, with normal hearing returning within 7–10 days.

Noise Potentiation of Cochleotoxicity. Synergism can also occur between ototoxic agents and noise (Bhattacharyya & Dayal, 1984; Boettcher, Henderson, Gratton, Danielson, & Byrne, 1987). Noise has been shown to potentiate the cochleotoxicity of salicylates (Chen & Aberdeen, 1980; McFadden & Plattsmier, 1983; Woodford, Henderson, & Hamernik, 1978), aminoglycosides (Brown, Brummett, Meikle, & Vernon, 1978; Dayal, Kokshanian, & Mitchell, 1971; Marques, Clark, & Hawkins, 1975), and cis-platin (Sharma & Edwards, 1983).

Return to Top

Incidence of Drug-Induced Cochleotoxicity

The actual frequency of cochleotoxicity associated with specific drugs is unclear because of inconsistencies in reported data. Aminoglycoside-induced cochleotoxicity has been reported to range from 0% (Powell, Thompson, & Luthe, 1983) to 63% (Tablan, Reyes, Rintelmann, & Lerner, 1984), and cisplatin cochleotoxicity from 3% (Forastiere, Takasugi, Baker, Wolf, & Kudla-Hatch, 1987) to 100% (Kopelman, Budnick, Sessions, Kramer, & Wong, 1988). Prospective studies using high-frequency audiometry have demonstrated that cochleotoxicity can occur initially or solely in the high-frequency range (9–20 kHz; Dreschler, Hulst, Tangs, & Urbanus, 1985; Dreschler, van der Hulst, Tange, & Urbanus, 1989; Fausti et al., 1993; Fausti et al., 1992b; Fausti et al., 1984a; Fausti et al., 1984b; van der Hulst, Dreschler, & Urbanus, 1988). However, most studies have restricted testing to the conventional-frequency range, potentially obscuring the actual risk of drug-induced loss of hearing sensitivity.

Variability in the criteria for defining cochleotoxicity also contributes to discrepant findings. Cochleotoxic change criteria have included increases in threshold at any single frequency of • 15 dB (Tablan et al., 1984; Thompson & Northern, 1981) or • 20 dB (Finley et al., 1982; van der Hulst et al., 1988); • 15 dB increase in threshold at two or more frequencies (Lerner, Schmitt, Seligsohn, & Matz, 1986; van der Hulst et al., 1988); average hearing decrease over a range of frequencies (Barr-Hamilton et al., 1991); or a combination of criteria (Pedersen, Jensen, Osterhammel, & Osterhammel, 1987). Pasic and Dobie (1991) advocate a data-based decision theory approach to recommending audiometric criteria. In many studies, no criteria are reported. Other variables affecting risk/frequency estimates include characteristics of study populations (gender, age, preexisting hearing loss, nature and severity of the illness under treatment), and variations in drug types, dosages, and treatment schedules.

Return to Top

Basic Principles of Cochleotoxicity Monitoring

The relationship between cochleotoxicity and drug administration parameters such as dosage, duration of treatment, and serum concentration is highly variable (Barza & Lauermann, 1978; Fausti et al., 1992b; Fausti et al., 1993; Schentag, 1980). An attending physician, therefore, cannot rely solely on dosage or serum concentrations to predict the risk of ototoxicity. The prospective assessment of hearing function remains the only reliable method for detecting the presence of cochleotoxicity prior to symptomatic hearing loss.

The higher incidence of cochleotoxicity seen in studies utilizing high-frequency (9–20 kHz) threshold evaluation (Aguilar-Markulis et al., 1981; Fausti et al., 1993; Fausti et al., 1992b; Helson, Okonkwo, Anton, & Cvitcovic, 1978) is a direct result of monitoring the frequency range in which cochleotoxic change is initially (and, sometimes, solely) detected. Thus, including high frequencies in monitoring programs will generally enable earlier identification of cochleotoxicity. Early detection of cochleotoxicity allows the physician to evaluate the potential for treatment alternatives that may prevent hearing loss in frequencies critical for speech communication, and alerts the audiologist to plan rehabilitation measures for patients in whom hearing loss cannot be avoided.

The available evidence indicates that high-frequency audiometry is the method of choice for the earliest detection of ototoxic hearing loss. Instrumentation is currently available to serve this purpose, and clinically acceptable intrasubject reliability using commercial high-frequency audiometers with either circumaural or insert earphones has been documented (Fausti, Frey, Henry, Knutsen, & Olson, 1990; Frank & Dreisbach, 1991; Frank, 1990; Laukli & Mair, 1985; Valente, Valente, & Goebel, 1992).

Return to Top

Elements of Monitoring

A basic cochleotoxicity monitoring program requires (a) specific criteria for identification of toxicity, (b) timely identification of at-risk patients, (c) pretreatment counseling regarding potential cochleotoxic effects, (d) valid baseline measures (pretreatment or early in treatment), (e) monitoring evaluations at sufficient intervals to document progression of hearing loss or fluctuation in sensitivity, and (f) follow-up evaluations to determine post-treatment effects.

Audiometric Criteria for Cochleotoxicity. Specific criteria for defining drug-induced hearing decrease are controversial (Simpson, Schwan, & Rintelmann, 1992). Here we have attempted to construct criteria by which most cases of true ototoxicity will be detected. These criteria are conservative, because the occasional false-positive identification is preferable to methods that may delay detection of the ototoxic process.

In a study of normal test-retest variability of audiometric thresholds, normal variability was reflected by independent shifts at random frequencies (Atherley, 1963). Thus, shifts at adjacent test frequencies indicate more systematic change and increase the likelihood of a true decrease in sensitivity (Dobie, 1983). Frequency averaging (i.e., calculating the average of thresholds across some frequency range) has been advocated for detecting decreased sensitivity, and the use of adjacent frequencies is equivalent to averaging over those frequencies. Another fundamental concept is that a decrease observed on repeated tests is a valid change (Royster & Royster, 1982). Thus, a shift relative to baseline that is seen at least twice is likely to represent a true shift and not normal variation.

Change in hearing sensitivity is always computed relative to baseline measures. Criteria to indicate hearing decrease during ototoxicity monitoring are defined here as (a) • 20 dB decrease at any one test frequency, (b) • 10 dB decrease at any two adjacent test frequencies, or (c) loss of response at three consecutive test frequencies where responses were previously obtained (the third criterion refers specifically to the highest frequencies tested, where earlier responses are obtained close to the limits of audiometric output and later responses cannot be obtained at the limits of the audiometer). Finally, change must be confirmed by repeat testing.

Patient Identification. Patients requiring monitoring are those whose treatment includes the administration of a therapeutic drug known or suspected to have cochleotoxic side effects. Once a patient is identified as being treated with a cochleotoxic drug, a monitoring program should be implemented in a timely manner. Access to a registry of hospitalized patients being treated with potentially cochleotoxic drugs is a critical component of a comprehensive monitoring program, and must often be developed in cooperation with hospital pharmacy personnel.

Pretreatment Counseling. Prior to treatment with cochleotoxic drugs, patients should be counseled regarding potential effects on the auditory system. During initial medical treatment counseling, the physician should include information regarding the risks and benefits of drug therapy. The audiologist should counsel patients on signs and symptoms of cochlear damage and potential effects on communication ability. Symptoms such as tinnitus, fullness, loss of balance, or changes in hearing sensitivity should be reviewed and the patient instructed to inform health care professionals if they occur. Potentiating effects such as exposure to noise during or following treatment should be discussed. If the patient lives or works in an environment with high noise levels, the possible synergistic effect of noise and cochleotoxic damage must be considered, and both the patient and family should be made aware of this increased risk.

Baseline Testing. The purpose of baseline testing is to document the status of hearing prior to treatment. At-risk individuals should receive baseline evaluations that are as complete as possible. Word discrimination should be included in the ideal scope of audiologic practice. The reliability of behavioral responses should be assessed during baseline by repeating selected portions of the evaluation. In addition, results of the first test following baseline should be evaluated for intertest reliability.

The optimal timing of baseline testing depends largely on the drug(s) the patient is receiving. For example, animals receiving large bolus doses of kanamycin do not show histologic evidence of cochleotoxicity until after 72 hours (Brummett, 1983; Brummett & Fox, 1982). Thus, in the absence of more precise data, baseline audiometric evaluation of patients receiving aminoglycosides should be done prior to or within 72 hours of the first treatment dose (Fausti et al., 1992b). Cisplatin can cause observable cochleotoxicity following a single course of treatment (Durrant, Rodgers, Meyers, & Johnson, 1990). Thus, it is important to obtain baseline measures prior to the first dose of cisplatin (Fausti et al., 1993).

Monitoring Schedule and Follow-Up Tests. Monitoring tests should be scheduled at intervals that will enable the earliest possible detection (within reason) of cochleotoxic effects. Immediate post-treatment testing is suggested to document auditory status at the end of drug treatment. Follow-up testing should be done at intervals appropriate to detect post-treatment cochleotoxicity, or to document recovery.

Return to Top

Cochleotoxicity Monitoring Procedures

Because of the variety of resources available in different medical/audiologic settings, suggested cochleotoxicity monitoring guidelines are fairly general. A more detailed monitoring protocol is outlined in the Appendix, which contains specific recommendations based on a series of cochleotoxicity investigations (Fausti, Frey, Rappaport, & Erickson, 1979; Fausti et al., 1993; Fausti et al., 1992b; Fausti et al., 1984a; Fausti et al., 1984b).

Patients receiving drug therapy can be placed into three general categories (based on their ability to perform behavioral measurement tasks): responsive, limited responsive, and unresponsive. Responsive patients are able to provide reliable behavioral responses to repeated comprehensive evaluations. Limited-responsive patients can provide reliable behavioral responses for only short periods of time because of illness, physical condition, or age-related factors. Unresponsive patients cannot provide reliable behavioral responses and can only be evaluated with objective measures that do not rely on responsiveness or attentiveness.

Return to Top

Responsive Patients

Baseline testing should be done as soon as possible after patient identification. For patients receiving antibiotics, baseline data should be obtained prior to or within 72 hours after initial drug administration. For patients receiving platinum-based chemotherapy, the baseline should be performed within 1 week prior to, and no later than 24 hours after the initial treatment. The baseline evaluation should be as comprehensive as possible, including immittance and speech tests.

Basic monitoring evaluations should include otoscopic examinations and air-conduction thresholds at (minimally) octave intervals from 0.25 to 8 kHz (omitting 0.25 kHz if testing is done at bedside). The schedule for monitoring evaluations depends on the type of ototoxic medication. Patients receiving antibiotics should be evaluated weekly. Patients receiving platinum derivatives should be evaluated within 24 hours prior to each course of treatment. The frequency of monitoring should be increased if decrease in hearing sensitivity is observed. If a decrease is noted, the patient should be retested within 24 hours to confirm the damage, and the physician should be informed as soon as possible of a validated change. Also, patients who complain of symptoms consistent with cochlear or vestibular damage must be seen immediately. This basic monitoring plan can be made more sensitive if the time and personnel are available. A more stringent plan includes monitoring every 2–3 days for patients receiving aminoglycosides, which could result in earlier identification of cochleotoxicity (Fausti et al., 1992b).

Any significant increase in air-conduction thresholds may warrant a complete audiologic evaluation. Because decrease in hearing could also result from middle ear involvement, immittance or bone-conduction testing should be done to differentiate sensorineural from conductive loss.

The patient should also be seen after cessation of treatment. Follow-up evaluations immediately after treatment and at 3 and 6 months post-treatment are optimal. If a decrease in hearing sensitivity is noted at any of these tests, weekly monitoring should be introduced until the hearing stabilizes. If hearing stabilizes at a level significantly worse than that shown on baseline testing, a complete differential audiologic battery is recommended.

Return to Top

Patients With Limited Responsiveness

Including test frequencies from 9 to 20 kHz provides maximum hearing sensitivity information with which to monitor hearing change. This can be a lengthy procedure, however, and may over-challenge an ill patient's ability to provide reliable responses. For patients whose responsiveness is limited by illness or age, monitoring procedures must accommodate the patient's decreased ability to provide reliable responses while obtaining the most essential information. Evaluation of fewer frequencies may be necessary for these patients. Testing should begin at the higher frequencies and progress downward as long as the patient continues to provide reliable responses. If test equipment has calibrated high-frequency (> 8 kHz) capability, testing should begin at the highest frequency appropriate to the patient. If, for example, the patient is young and has no history of ear pathology, the appropriate initial frequency would be higher than that for an older patient with a history of noise exposure or specific ear pathology. If only partial information can be obtained initially, complete data should be obtained if and when the patient is capable.

Specific protocols have been designed for patients with limited responsiveness, based on auditory monitoring data from a large group of hospitalized patients receiving cisplatin and aminoglycosides (Fausti et al., 1993; Fausti et al., 1992b). These studies have identified a limited range of frequencies relative to an individual patient's hearing threshold configuration. Monitoring only these frequencies has demonstrated a high degree of sensitivity to early decrements in hearing as a consequence of drug therapy.

Return to Top

Unresponsive Patients

In nonresponsive patients, objective hearing measures (e.g., auditory evoked potentials and evoked otoacoustic emissions) may be the only means to obtain auditory information. Although objective procedures provide only gross information on hearing sensitivity, they are, nonetheless, capable of detecting ototoxic hearing loss (Bernard, Pechere, & Herbert, 1980; Guerit, Mahieu, Houben-Giurgea, & Herbay 1981; Hall, Herndon, Gary, & Winkler, 1986; Piek, Lumenta, & Bock, 1985).

Hearing evaluation in patients unable to provide reliable behavioral responses is a complex issue. Aside from the question of which objective assessment technique to use, there is the medicolegal concern of informed consent. For unresponsive patients, proxy consent must be obtained according to the laws of the state and the policies and procedures of the specific institution.

Three objective evaluation procedures have potential for ototoxicity monitoring of unresponsive patients: otoacoustic emissions (OAE), electrocochleography (ECochG), and auditory brainstem response (ABR). These techniques are in various stages of development for use as objective ototoxicity monitoring tools. Thus, specific guidelines for application of these procedures to ototoxicity monitoring cannot be recommended at this time. Under conditions when no behavioral response is available, however, use of objective measures is encouraged. Responses obtained in this manner at least document the gross responsiveness of the auditory system. Repeated testing can be informative regarding changes during treatment. At a minimum, the absence of a previously obtained response indicates that gross auditory function has been reduced or lost.

OAE. Three types of OAE measurements have received concentrated attention: spontaneous, transiently evoked, and distortion product. OAE assessment is specifically sensitive to the status of outer hair cells in the cochlea and is a relatively efficient objective test. It has been used to assess cochlear function in patients receiving cisplatin with promising results (Beck, Maurer, Welkoborsky, & Mann, 1992; Plinkert & Krober, 1991). Although OAE testing presents a new and exciting tool for cochleotoxicity monitoring, its application has not been evaluated sufficiently to enable formulation of specific guidelines.

ECochG. This evoked potential technique has been used for a number of years to evaluate cochlear and neural responses. The most sensitive ECochG technique requires transtympanic placement of the electrode and is, therefore, more invasive than OAE or ABR. Tympanic membrane or ear canal placements are less invasive, but also provide less sensitive threshold estimates. Furthermore, the use of these special electrode placements may not be suitable for all patients. In any configuration, ECochG requires a significant amount of time to acquire frequency-specific information. This technique therefore is not appropriate for routine objective auditory monitoring.

ABR. ABR is subject to the same limitations as ECochG with respect to length of testing and frequency specificity. The use of acoustic clicks as stimuli limits response information to frequencies between 1 and 4 kHz, and thus decreases its effectiveness in cochleotoxicity monitoring. Studies using high-frequency tone-burst stimuli to provide high-frequency-specific response information, however, have shown that this ABR technique holds promise as an objective monitoring tool for early detection of ototoxicity (Fausti, Frey, Henry, Olson, & Schaffer, 1992a). Recent advances in multiple-stimulus ABR (Hamill, Hussung, & Sammeth, 1991; Hammil, Yanez, Collier, & Lionbarger, 1992; Hoke, Pantev, Ansa, Lutkenhoner, & Herrmann, 1991) may shorten the test time when such developments are incorporated into clinical instrumentation.

Return to Top

Testing Environments

Audiologic evaluation ideally should be conducted in a sound-treated booth. When a patient is unable to leave the ward environment, however, or if sound booth testing cannot be performed for any reason, audiometric testing may be conducted at bedside. Generally, the higher frequencies are less vulnerable to masking by ambient noise than lower frequencies. Thompson and Northern (1981), in advocating bedside monitoring with portable audiometers in the conventional-frequency range, concluded that ambient noise is a problem primarily at low frequencies and not a significant concern at frequencies from 2 to 8 kHz (and, presumably, above 8 kHz). Reliability of threshold testing at 8 kHz and above has been demonstrated in bedside testing of normal-hearing subjects (Valente et al., 1992), and those with sensorineural hearing loss (Sinks & Goebel, 1993). Thus, masking effects of ambient noise appear to be less problematic than once thought.

Although there are no standards for maximum permissible ambient noise levels when testing high-frequency thresholds, ambient noise levels in the test environment should be monitored with a sound-level meter and recorded when conducting bedside audiometric testing (Fausti et al., 1992b; Valente et al., 1992). Sound level measurements are used in assessing response reliability by evaluating differences in ambient noise between sessions. The type of ambient noise analysis will depend on the sound-measuring equipment available. One-third octave band analysis is the ideal, but one-octave band analysis is sufficient (ANSI, 1991).

Return to Top

Management of Children

Prior to therapeutic treatment with potentially ototoxic drugs in children, parents should be counseled regarding the potential for ototoxicity and consequent hearing loss. Information obtained from parents at the time of baseline testing should include the child's history of speech/language and motor development and a discussion of test findings. Parents should be alerted to clinical signs and symptoms indicative of cochlear/vestibular function change that they should report to their physician.

Young children may be at greater risk for ototoxicity than similarly treated adult patients (Schell et al., 1989). Pasic and Dobie (1991) observed hearing loss in 77% of children receiving cisplatin. Even a minimal hearing loss is debilitating for young children acquiring speech and language skills.

Long-term audiologic follow-up is important to determine when hearing loss is stable or progressive. Follow-up testing should occur immediately, and at 3 months, 6 months, and 1 year after cessation of treatment. Frequent monitoring and early identification permits early intervention as appropriate.

Return to Top

Post-Treatment Aural Rehabilitation

If an ototoxic hearing loss results in communication deficit, the audiologist is ethically bound to begin, or recommend, aural rehabilitation (including amplification, assistive listening devices, speechreading, etc.). Intervention should begin as soon as possible after hearing loss has been identified.

Appendix — Suggested Procedures for Cochleotoxicity Monitoring